Brooklyn is a highly populated urban setting with a high density of food stores, of which supermarkets comprise a small proportion – 4% on average. The food retail environment in Brooklyn is dynamic: there was an overall increase in total stores and supermarkets throughout Brooklyn, but there were also many store closures during the five-year period. These closures were offset by food store replacements as well as store openings in new locations (including 23 new supermarkets). Stability of the food retail environment varied by wealth and racial/ethnic make-up of census tracts. Differences were observed in the density of stores, proportion of supermarkets, and individual store stability (proportion of stores that were open for the entire five-year period). Although there were fewer stores per 10,000 people in the highest-income and predominantly white areas, store stability was higher in these areas. Individual supermarkets were more stable than the general food retail environment in all race and income areas, yet supermarket stability followed a similar pattern – fewer supermarkets closed in higher-income areas. While the proportion of all 2011 stores in new locations was similar throughout race areas and income areas, new supermarket locations in the lowest-income areas were disproportionately high compared with other income groups. However, proportion of all stores that were supermarkets in 2011 was still higher in the highest-wealth areas.
The reasons for the observed differences between neighborhoods and the fluctuations in the food retail environment cannot be determined using the data provided. These data represent a specific period, which encompassed an economic recession in the United States . These economic factors may have influenced the stability of the food retail environment. For example, it is possible that a decrease in real estate value influenced the number of store openings. However, since lower-income neighborhoods experienced greater instability in the food marketplace (more store closings and openings), one could surmise that businesses in those neighborhoods had fewer resources to help them withstand economic downturns than those in the highest wealth areas, and many could not survive for long. Although the trends observed in Brooklyn during the study period may not be generalizable to any other five-year period, the findings are important for considering the impact of food retail fluctuations on the shopping patterns of affected residents.
The most salient data presented here is the amount of fluctuation in neighborhood food retail, and the varying degrees of stability between areas by wealth and racial characteristics. The findings highlight the need to understand how consumers utilize adaptive behaviors, or modify their food shopping patterns, to navigate a frequently changing environment. In neighborhoods where stores often change ownership or close permanently, residents may change their shopping behaviors in any number of ways. They may spend more time adjusting their shopping routines than those in more stable food environments. The adjustments necessary are not only geographic (finding the next best location to shop), but also economic (adjusting to different prices at the new shopping location), and dietary (adjusting to varying selections at other stores). Where closed food stores are replaced, access is restricted in the interim between a closing and an opening. Permanent closures of supermarkets in particular disrupt residents’ access to food [11–13]. While permanent supermarket closures were much lower than that of total stores, the closing of just one supermarket – where the next nearest is several blocks away or in a different neighborhood – might mean having to find transportation and additional time to travel farther to shop. Further, older and less ambulatory residents affected by a supermarket closing may begin relying on smaller stores – which likely offer a comparably limited range of goods at higher prices. In this study, all area types except highest-income areas were affected by permanent closures of supermarkets.
Researchers, policymakers, and others might consider how they can utilize fluctuations in the food retail environment to further the goal of improving access to healthy foods. For example, these data demonstrate that non-supermarket food stores are the most prevalent and have the highest turnover. The turnover is particularly pronounced in the lowest-income study areas and in predominantly non-white and mixed population areas. On one hand, these fluctuations may create unpredictable shopping environments for neighborhood residents. On the other, fluctuation – combined with the high prevalence of non-supermarket stores – may indicate tremendous opportunity for public health policy to work with store owners and community partners to influence the availability of healthy foods in these neighborhoods, as has been suggested by others [14–17]. When a food store changes hands or simply upgrades its image, the owner may be open to new ideas. Also, such changes may be promising to consumers who are looking to broaden their options. Therefore, fluctuations in the food retail environment may provide an open window to reach both store owners and residents.
This analysis has some limitations, which should be considered when interpreting the findings. First, for race and income analyses we used Census 2000 data and therefore were not able to consider demographic changes that occurred in these neighborhoods during the study period. The reason for this was that by the 2010 census, some census tract boundaries had changed. Second, we used government data to identify food stores. While these data are updated regularly, they do not include store type categories, which limited our analysis to total stores and stores validly identified as supermarkets. Third, this analysis is limited to the number of stores where food is available, without investigating the availability of certain food items at those stores, or their prices and quality. Several studies have investigated these factors and found that 1) smaller stores and low-income neighborhoods are associated with poorer quality of produce [18, 19] and 2) food stores located in low-income and neighborhoods of color have lower availability of food items considered healthier, as well as higher prices for these items [5, 9, 20]. Other work has described the impact of food quality on purchasing behavior [21–23]. Finally, since an urban foodscape differs from other environments in terms of the density of population and stores, our findings may not be generalizable to other areas of the United States or internationally.
Despite these limitations, the data presented are to our knowledge the first to demonstrate the dynamic nature of food retail environments, the data represent, the entirety of one urban setting and not a sample, enabling a comparison of neighborhoods by race and income that illuminates disparities for some residents. As in other urban settings, the dense population of Brooklyn (nearly 2.5 million people) underscores the need to understand how residents are affected by fluctuations in their food environments. In Brooklyn, there are over one million residents in predominantly non-white neighborhoods, and 650,000 residents in the lowest-income neighborhoods – areas that these data show are the least stable food retail environments.